Social and Behavioral Determinants of Sleep

Abstract

There has been an epidemic of sleep deficiency over the past few decades. Previous research has linked suboptimal sleep to various adverse outcomes (e.g., higher morbidity and mortality), which warrant efforts to identify factors that predispose people to sleep deficiency. This three-chapter dissertation aims to examine multiple social and behavioral determinants of sleep among various populations. The Background Chapter presents an overview of sleep deficiency. Chapter 1 investigates predictors of persistent sleep problems among elderly disaster survivors. Data came from a prospective cohort that suffered the 2011 Great East Japan earthquake and tsunami between baseline and follow-up surveys (N = 3,547). Poisson regression with robust error variance models were used. We found that material aspects of disaster damage (financial hardship, home destruction, health care disruption) were associated with persistent sleep problems. In contrast, elderly survivors appeared to have recovered from loss of loved ones three years after the disaster. Having pre-disaster instrumental support reduced the risks of sleep problems. Findings could inform targeted disaster recovery efforts for groups with greatest need. Chapter 2 explores the relationships between procrastination and sleep among US adolescents and young adults. Multiple linear regression and Poisson regression with robust error variance models were used to analyze data from the Healthy Minds Study (N = 8,742). Results showed that higher levels of procrastination were associated with worse sleep among US undergraduate and graduate students. Interventions to prevent and manage procrastination might help students to improve their sleep health. Chapter 3 examines how adolescents’ positions relative to cohesive friendship groups and popularity among peers were associated with their sleep behaviors. Data came from two schools in the National Longitudinal Study of Adolescent to Adult Health (N = 2,553). Multiple linear regression and Poisson regression with robust error variance models were used. We found that the absence of membership in a cohesive group was associated with suboptimal sleep. Also, there seemed to be a burden of being popular when it came to sleep health. Results suggested that a social network perspective might inform interventions to improve adolescents’ sleep outcomes.